Power distribution network operation regulation method and system

Abstract

The application discloses a power distribution network operation regulation method and system, which is applied to the field of power distribution network operation regulation and comprises the following steps: dividing historical topology change data of a power distribution network to obtain type change data of each topology type label; clustering the type change data based on historical regulation strategies corresponding to each topology type label to obtain a strategy clustering set with a strategy label; determining a related feature set of the strategy label based on regulation evaluation values corresponding to the strategy clustering set; and regulating the power distribution network based on the historical regulation strategies and the related feature set. The application determines regulation strategies under each topology change and related features of strategy adaptation, and realizes efficient regulation of the power distribution network in the application scene with variable power distribution network topology.

Description

Technical Field

[0001] This invention relates to the field of power distribution network operation and control technology, and in particular to a power distribution network operation and control method and system. Background Technology

[0002] As a crucial component of the power system, the operational efficiency of the distribution network directly impacts the stability of power supply. During distribution network operation, frequent topology changes are often triggered by factors such as distributed generation and dynamic load fluctuations. Existing distribution network control methods struggle to efficiently adapt to these topology shifts and the resulting changes in operational scenarios. In other words, current distribution network control strategies only accommodate a limited range of distribution network topology states and cannot effectively regulate distribution network operation under diverse topologies. Summary of the Invention

[0003] This invention provides a distribution network operation control method and system to solve the technical problem that existing distribution network operation control methods cannot efficiently control distribution networks with variable topologies.

[0004] To address the aforementioned technical problems, embodiments of the present invention provide a method for controlling and regulating the operation of a power distribution network, comprising: The historical topology change data of the distribution network is divided to obtain the type change data of each topology type label; Based on the historical control strategies corresponding to each of the topology type labels, the type change data are clustered to obtain a set of strategy clusters with strategy labels. Based on the regulation evaluation value corresponding to the strategy cluster set, determine the relevant feature set of the strategy label; The power distribution network is regulated based on the historical control strategies and the relevant feature set.

[0005] As one preferred embodiment, the step of dividing the historical topology change data of the distribution network to obtain type change data for each topology type label includes: Labels are extracted from the topology change data in the historical topology change data to obtain the label values ​​of each branch. Based on the label values ​​of each branch label, an initial adjacency matrix and a reconstructed adjacency matrix are constructed; Feature extraction is performed on the topological difference graph determined based on the initial adjacency matrix and the reconstructed adjacency matrix to obtain topological feature data; Based on the topological feature data, the historical topological change data is divided to obtain the type change data of each topological type label.

[0006] As one preferred embodiment, the step of clustering the type change data based on the historical control strategies corresponding to each of the topology type labels to obtain a strategy cluster set with strategy labels includes: Extract the historical strategies corresponding to each topology change from the type change data of each topology type label to obtain the historical control strategies corresponding to each topology type label; Based on the control logic and control method of the historical control strategy, the type change data are clustered to obtain the strategy cluster set corresponding to each topology type label; Based on the common characteristics of the strategies in the strategy cluster set, the strategy labels of the strategy cluster set are determined.

[0007] As one preferred embodiment, determining the relevant feature set of the strategy label based on the regulation evaluation value corresponding to the strategy cluster set includes: The regulation evaluation values ​​corresponding to each topological change in the strategy cluster set are sorted to obtain the effect evaluation sequence of the strategy label; Based on the effect evaluation sequence, the mean of the regulation evaluation distribution is obtained, and the comparison result between the mean of the regulation evaluation distribution and the mean threshold is determined. Based on the effect evaluation sequence and the comparison results, the range of the regulatory effect label of the strategy label is determined; Based on the regulation evaluation value and the regulation effect label range, the relevant feature set of the strategy label is determined.

[0008] As one preferred embodiment, the step of determining the relevant feature set of the strategy label based on the regulation evaluation value and the regulation effect label range includes: Based on the regulation evaluation value and the regulation effect label range, determine the regulation effect data of the strategy label; Determine the change characteristics corresponding to each topological change in the regulation effect data; A correlation analysis is performed on the change characteristics and the regulation evaluation value to obtain the relevant feature set of the strategy label.

[0009] As a preferred embodiment, the regulation of the distribution network based on the historical regulation strategy and the relevant feature set includes: Using the change features as model input, the historical control strategies and control evaluation values ​​corresponding to each topological change in the control effect data as model output, and the relevant feature set as model constraints, a migration model for the operating strategy of each topological type label is constructed. The operational strategy migration models of each of the aforementioned topology type labels are integrated to obtain the power grid strategy migration model; The power distribution network is regulated based on the power grid strategy migration model.

[0010] Another embodiment of the present invention provides a power distribution network operation and control system, comprising: The type change data determination module is used to divide the historical topology change data of the distribution network and obtain the type change data of each topology type label; The strategy cluster set determination module is used to cluster the type change data based on the historical control strategies corresponding to each topology type label to obtain a strategy cluster set with strategy labels. The relevant feature set determination module is used to determine the relevant feature set of the strategy label based on the regulation evaluation value corresponding to the strategy cluster set; The distribution network operation and control module is used to control the distribution network based on the historical control strategy and the relevant feature set.

[0011] As one preferred embodiment, the type change data determination module includes: The label extraction unit is used to extract labels from the topology change data in the historical topology change data to obtain the label values ​​of each branch label. The adjacency matrix construction unit is used to construct an initial adjacency matrix and reconstruct the adjacency matrix based on the label values ​​of each branch label; The topology feature data determination unit is used to extract features from the topology difference graph determined based on the initial adjacency matrix and the reconstructed adjacency matrix to obtain topology feature data. The type change data determination unit is used to divide the historical topology change data based on the topology feature data to obtain the type change data of each topology type label.

[0012] As one preferred embodiment, the strategy clustering set determination module includes: The historical control strategy determination unit is used to extract the historical strategy corresponding to each topology change in the type change data of each topology type label, and obtain the historical control strategy corresponding to each topology type label. The data clustering unit is used to cluster the type change data based on the control logic and control method of the historical control strategy to obtain the strategy cluster set corresponding to each topology type label; The strategy label determination unit is used to determine the strategy label of the strategy cluster set based on the common characteristics of the strategies in the strategy cluster set.

[0013] As one preferred embodiment, the relevant feature set determination module includes: The effect evaluation sequence determination unit is used to sort the control evaluation values ​​corresponding to each topological change in the strategy cluster set to obtain the effect evaluation sequence of the strategy label; The comparison result determination unit is used to obtain the mean of the regulation evaluation distribution based on the effect evaluation sequence, and to determine the comparison result between the mean of the regulation evaluation distribution and the mean threshold. The regulation effect label interval determination unit is used to determine the regulation effect label interval of the strategy label based on the effect evaluation sequence and the comparison results; The relevant feature set determination unit is used to determine the relevant feature set of the strategy label based on the regulation evaluation value and the regulation effect label range.

[0014] This invention provides a method and system for controlling the operation of a distribution network. The method involves dividing historical topology change data of the distribution network according to topology type labels, resulting in type change data under multiple topology type labels. Then, based on the historical control strategies corresponding to each topology type label, cluster analysis is performed on the type change data under each topology type label to obtain a strategy cluster set and its label. Next, using the control evaluation value corresponding to the strategy cluster set, relevant features for strategy label adaptation are selected. Finally, based on the historical control strategies and relevant policy adaptation features, the distribution network under each topology change is controlled. This invention achieves efficient control of the distribution network in application scenarios with highly variable distribution network topologies by determining the control strategies and relevant policy adaptation features under each topology change. Attached Figure Description

[0015] Figure 1 This is one of the flowcharts of the power distribution network operation and control method provided by the present invention; Figure 2 This is the second flowchart of the power distribution network operation and control method provided by the present invention; Figure 3 This is a schematic diagram of the structure of the power distribution network operation and control system provided by the present invention.

[0016] Figure label: Among them, 301 is the module for determining type change data; 302 is the module for determining strategy clustering set; 303 is the module for determining relevant feature set; and 304 is the module for distribution network operation and control. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0018] In the description of this application, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," "third," etc., may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0019] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal communication between two components. The terms "vertical," "horizontal," "left," "right," "upper," "lower," and similar expressions used herein are for illustrative purposes only and do not indicate or imply that the device or component referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as limiting the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0020] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this specification is for the purpose of describing specific embodiments only and is not intended to limit the invention. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0021] See Figure 1 , Figure 1 This is a flowchart illustrating an embodiment of the power distribution network operation and control method provided by the present invention, as shown below. Figure 1 As shown, this embodiment includes steps 100 to 400, and the specific steps are as follows: Step 100: Divide the historical topology change data of the distribution network to obtain the type change data of each topology type label; Specifically, historical change data (i.e., historical topology change data in this embodiment) is collected for each topology change during the historical operation of the distribution network. This historical change data includes at least topology structure change data, source-load change data, operational status data, topology type labels corresponding to each topology change, historical environmental data, historical control strategies corresponding to each topology change, and control evaluation values ​​corresponding to each topology change. Specifically, the topology structure change data includes the initial topology, branch labels and their values ​​for each branch in the initial topology, the reconstructed topology, and branch labels and their values ​​for each branch in the reconstructed topology; the source-load change data includes source-load characteristics and their characteristic values; the operational status data includes status characteristics and their characteristic values; and the historical environmental data includes environmental characteristics and their characteristic values.

[0022] The collected historical topology change data provides complete information covering different topology change scenarios. Each topology change is then processed as a unit, and the core information related to the topology change is extracted from the historical topology change data corresponding to each topology change. By analyzing the structural differences before and after the topology change, the key features that can characterize the core attributes of the topology change are extracted from the core information, which are the topology feature data corresponding to each topology change.

[0023] Based on the topology type label corresponding to each topology change, the historical topology change data of the distribution network is divided to obtain the type change data of each topology type label. The type change data of each topology type label includes historical topology change data of multiple topology changes of the same topology type label and topology feature data.

[0024] Step 200: Cluster the type change data based on the historical control strategies corresponding to each topology type label to obtain a strategy cluster set with strategy labels; Specifically, for the type change data of each topology type label, the historical control strategies for each topology change are extracted from the type change data of each topology type label. Historical control strategies refer to the historical control schemes adopted to adapt to the new topology after a topology change occurs. Based on the similarity of historical control strategies (e.g., similar control logic and similar control methods), type change data with similar control strategies in the type change data of each topology type label are grouped together to form multiple strategy cluster sets under each topology type label. A unique strategy label is assigned to each strategy cluster set.

[0025] Step 300: Based on the regulation evaluation value corresponding to the strategy cluster set, determine the relevant feature set of the strategy label; Specifically, the analysis is based on the strategy cluster sets corresponding to each strategy label under each topology type label. The regulation evaluation value (from historical topology change data) for each topology change within each strategy cluster set is extracted. The regulation evaluation value refers to the quantitative evaluation index value of the effect after the implementation of historical regulation strategies. Based on the regulation evaluation value, the uniform distribution value of all regulation evaluation values ​​for each strategy label under each topology type label is calculated. The obtained uniform distribution value is compared with a preset threshold. If the uniform distribution value is greater than the preset threshold, a uniform regulation effect label interval containing three categories of regulation effect labels (e.g., excellent, medium, and poor) is obtained based on the distribution range of regulation evaluation values ​​for all strategy labels (with uniform distribution values ​​greater than the preset threshold) under each topology type label. If the uniform distribution value is less than or equal to the preset threshold, a non-uniform regulation effect label interval containing three categories of regulation effect labels is obtained based on the same effect label classification standard, using the same distribution range of regulation evaluation values ​​for all strategy labels (with uniform distribution values ​​less than or equal to the preset threshold) under each topology type label.

[0026] Under each topology type label, the uniform or non-uniform control effect label interval corresponding to each strategy label is analyzed to determine the correlation between different features and the control evaluation values ​​corresponding to various control effect labels. This yields features that have a positive impact on the control effect labels and features that have a negative impact on the control effect labels, forming the relevant feature set in this embodiment.

[0027] Step 400: Based on the historical control strategy and the relevant feature set, control the distribution network.

[0028] Specifically, the process of constructing the operational strategy transfer model for each topology type label is as follows: The source load change data, operational status data, and topology feature data from the type change data of each topology type label are used as model inputs. The historical control strategies and control evaluation values ​​from the type change data of each topology type label are used as model outputs. The set of positive and negative correlation features is used as the core constraint condition for model construction. By constructing the operational strategy transfer model for each topology type label, a precise mapping from features to strategies is established, enabling the operational strategy transfer model to analyze the control strategy adaptation logic of that topology type label. When a new topology scenario is generated, the core feature data of the new topology scenario are input into the operational strategy transfer model for each topology type label to obtain the adapted control strategy and control prediction effect output by the model.

[0029] After the operation strategy migration model of each topology type label is running stably, the operation strategy migration models corresponding to all topology type labels are then globally integrated to form a unified model that can adapt to all topology scenarios. This unified model is then used to perform strategy control on each topology scenario of the distribution network.

[0030] In this embodiment, historical topology change data of the distribution network is divided according to topology type labels, resulting in type change data under multiple topology type labels. Then, based on the historical control strategies corresponding to each topology type label, cluster analysis is performed on the type change data under each topology type label to obtain a strategy cluster set and its strategy label. Next, using the control evaluation value corresponding to the strategy cluster set, relevant features for strategy label adaptation are screened. Finally, based on the historical control strategies and relevant policy adaptation features, the distribution network under each topology change is controlled. This invention achieves efficient control of the distribution network in application scenarios with highly variable distribution network topologies by determining the control strategies and relevant policy adaptation features under each topology change.

[0031] See Figure 2 , Figure 2 This is a flowchart illustrating another embodiment of the power distribution network operation and control method provided by the present invention, as shown below. Figure 2 As shown, this embodiment includes steps 110 to 140, and the specific steps are as follows: Step 110: Extract labels from the topology change data in the historical topology change data to obtain the label values ​​of each branch label; Step 120: Based on the label values ​​of each branch label, construct the initial adjacency matrix and reconstruct the adjacency matrix; Step 130: Extract features from the topological difference graph determined based on the initial adjacency matrix and the reconstructed adjacency matrix to obtain topological feature data; Step 140: Divide the historical topology change data based on the topology feature data to obtain the type change data of each topology type label.

[0032] Specifically, historical topology change data for each topology change is obtained from the historical operation data of the distribution network. By analyzing the structural differences before and after each topology change, key features characterizing the core attributes of the topology change are extracted to obtain the topology feature data corresponding to each topology change. This topology feature data can be used to segment historical topology change data. When segmenting type change data for each topology type label, the topology feature data is used as the basis for data segmentation. For example, type change data for the same topology type label have consistent structural features.

[0033] The process of extracting topology feature data is as follows: Historical topology change data for each topology change includes the initial topology and the reconstructed topology. The initial topology is the original network structure of the distribution network before the topology change, and the reconstructed topology is the new network structure formed after the topology change. Each branch in both the initial and reconstructed topologies has a distinguishing branch label. The labels of all branches in both structures are extracted to obtain initial branch labels and reconstructed branch labels, each with a corresponding label value. For example, branch labels can include main feeders, first-level branches, second-level branches, and user branches, with corresponding label values ​​of 1, 2, 3, and 4, respectively. Based on the branch labels and their values ​​corresponding to the initial topology, an initial adjacency matrix is ​​constructed. Similarly, based on the branch labels and their values ​​corresponding to the reconstructed topology, a reconstructed adjacency matrix is ​​constructed. The topology difference graph is determined using the initial and reconstructed adjacency matrices corresponding to each historical topology change. Feature extraction is performed on the topology difference graph to obtain the topology feature data for each historical topology change. Topological feature data includes multiple topological features and the corresponding topological feature values ​​for each topological feature. Topological feature values ​​can be the number of branch breaks, the number of branch closures, and the number of branch label changes corresponding to each label value, etc.

[0034] This embodiment extracts topological feature data from each topological change to achieve a quantitative characterization of the structure before and after the topological change, providing data support for subsequent content of this invention.

[0035] In another embodiment of the power distribution network operation control method provided by the present invention, step 200 specifically includes: Step 210: Extract the historical strategies corresponding to each topology change in the type change data of each topology type label to obtain the historical control strategies corresponding to each topology type label; Step 220: Based on the control logic and control method of the historical control strategy, cluster the type change data to obtain the strategy cluster set corresponding to each topology type label; Step 230: Determine the policy labels of the policy cluster set based on the common policy features of the policy cluster set.

[0036] Specifically, historical control strategies are extracted from historical topology change data of multiple topology changes under the same topology type label, resulting in all historical control strategies corresponding to each topology type label. Based on these historical control strategies, cluster analysis is performed on the type change data for each topology type label. The clustering criterion is to group type change data with similar control logic and methods together, resulting in multiple data subsets (i.e., strategy cluster sets in this embodiment). According to the differences in control strategies, multiple non-overlapping strategy cluster sets are generated under each topology type label. Then, the core common features of all historical control strategies in each strategy cluster set are extracted to obtain the strategy label for each strategy cluster set. In summary, each strategy cluster set contains historical topology change data and its topology feature data for one or more topology changes. Strategy labels can include conservative voltage boosting, aggressive loss reduction, reactive power balancing, and energy storage peak shaving, etc.

[0037] This embodiment uses historical control strategies to cluster the type change data of each topology type label, achieving accurate aggregation of similar control strategies under the same topology type, and providing accurate data support for subsequent solutions.

[0038] In another embodiment of the power distribution network operation control method provided by the present invention, the above steps specifically include: Step 310: Sort the regulation evaluation values ​​corresponding to each topological change in the strategy cluster set to obtain the effect evaluation sequence of the strategy label; Step 320: Based on the effect evaluation sequence, obtain the mean of the regulation evaluation distribution, and determine the comparison result between the mean of the regulation evaluation distribution and the mean threshold. Step 330: Based on the effect evaluation sequence and the comparison results, determine the regulatory effect label range of the strategy label; Step 340: Based on the regulation evaluation value and the regulation effect label range, determine the relevant feature set of the strategy label.

[0039] Specifically, all regulation evaluation values ​​for each strategy label under each topology type label are sorted to obtain the effect evaluation sequence for each strategy label under each topology type label. Based on the effect evaluation sequence for each strategy label under each topology type label, the process of calculating the mean of the regulation evaluation distribution for each strategy label under each topology type label is shown in Formulas 1 to 4. Wherein, The interval for evaluating the regulatory effect of the policy clustering set of the i-th topology type label and the j-th policy label; This represents the k-th regulation evaluation value in the effect evaluation sequence of the j-th strategy label of the i-th topology type label; The number of segments in the strategy cluster set of the i-th topology type label and the j-th strategy label; The number of regulatory evaluation values ​​in the effect evaluation sequence of the i-th topology type label and the j-th strategy label; The a-th segment of the effect evaluation sequence for the j-th strategy label of the i-th topology type label; The number of regulation evaluation values ​​in the a-th segment of the effect evaluation sequence of the j-th strategy label of the i-th topology type label; To adjust the parameters evenly.

[0040] (1) (2) (3) (4) According to Formulas 1 to 4, the number of regulation evaluation values ​​in all segments of the effect evaluation sequence of the i-th topology type label and the j-th strategy label is close to the average number of segments. When the distribution of control evaluation values ​​is uniform, the distribution of control evaluation values ​​is uneven when there is a significant peak or sparsity in the number of control evaluation values ​​in all segments of the effect evaluation sequence of the i-th topology type label and the j-th strategy label. The larger the value, the more sensitive it is to uniform distribution, and the stronger the penalty for uneven distribution. The smaller the value, the weaker the penalty for unevenness. In this embodiment, the mean threshold is used to distinguish whether all the regulation evaluation values ​​of each policy label for each topology type label are relatively uniform.

[0041] The mean value of the regulation evaluation distribution of each strategy label for each topology type is compared with the mean threshold. Based on the effect evaluation sequence of all strategy labels (where the mean value of the regulation evaluation distribution is greater than the preset threshold) for each topology type, the uniform regulation effect label interval for each strategy label for each topology type is calculated. The regulation effect labels are categorized as excellent, medium, and poor, as shown in Formulas 5 to 8. The uniform control effect label interval of the policy clustering set of the i-th topology type label and the j-th policy label; The uniform adjustment lengths of the first, second, and third sub-intervals of the policy clustering set for the policy clustering set ... These are the first, second, and third sub-intervals for evaluating the regulatory effect of the policy clustering set for the i-th topology type label and the j-th policy label, respectively. This is the range adjustment factor; It represents the relative policy control value for all sub-topology changes in the policy cluster set of the i-th topology type label and the j-th policy label; To prevent zero parameters; It is the average regulation evaluation value of all sub-topology changes in the strategy cluster set of the i-th topology type label and the j-th strategy label; The number of strategy cluster sets for the i-th topology type label; The effect evaluation sequence of the i-th topology type label and the j-th policy label. One regulatory assessment value; This indicates rounding up to the nearest integer.

[0042] (5) (6) (7) (8) According to formulas 5 to 8, The larger the value, the better the regulation effect of the regulation of all topological changes in the policy cluster set of the corresponding topological type label.

[0043] Based on the effect evaluation sequences of all strategy labels (with the mean of the regulation evaluation distribution less than or equal to the preset threshold) for each topology type label, the non-uniform regulation effect label intervals for each strategy label of each topology type label are calculated. The regulation effect labels are categorized as excellent, medium, and poor, as shown in Formulas 9 to 12. The label interval of the non-uniform regulation effect of the policy clustering set of the i-th topology type label and the j-th policy label; and Let represent the first and second partition points of the effect evaluation sequence of the j-th policy label for the i-th topology type label, respectively; The valley point of the a-th segment in the effect evaluation sequence of the i-th topology type label and the j-th strategy label; Let be the set of valleys in the effect evaluation sequence of the i-th topology type label and the j-th policy label; and .

[0044] (9) (10) (11) (12) Based on the above-obtained regulation evaluation values, uniform regulation effect label intervals, and non-uniform regulation effect label intervals, the set of positive and negative correlation features for each strategy label of each topology type is determined.

[0045] This embodiment achieves deep adaptation between features and scenarios by determining the set of positive and negative correlation features of each strategy label for each topology type label.

[0046] In another embodiment of the power distribution network operation control method provided by the present invention, step 340 specifically includes: Step 341: Based on the regulation evaluation value and the regulation effect label range, determine the regulation effect data of the strategy label; Step 342: Determine the change characteristics corresponding to each topological change in the regulation effect data; Step 343: Perform correlation analysis on the change characteristics and the regulation evaluation value to obtain the relevant feature set of the strategy label.

[0047] Specifically, based on the regulation assessment value, the uniform regulation effect label interval, and the non-uniform regulation effect label interval, the specific process for determining the set of positive and negative correlation features for each strategy label of each topology type is as follows: Pearson correlation analysis is performed on each source load feature in the source load change data of each strategy label's regulation effect data and the regulation assessment value to obtain the source load correlation coefficient between each source load feature and the regulation assessment value. Similarly, the state correlation coefficient between each state feature and the regulation assessment value, the environmental correlation coefficient between each environmental feature and the regulation assessment value, and the topological correlation coefficient between each topological feature and the regulation assessment value can be obtained.

[0048] Based on the correlation coefficients obtained above, the first, second, and third correlation vectors for each strategy label of each topology type are determined. All source-load correlation coefficients, state correlation coefficients, environmental correlation coefficients, and topology correlation coefficients corresponding to the control effect data with the control effect label "excellent" under each strategy label are integrated to obtain the first and second correlation vectors, respectively. Similarly, all source-load correlation coefficients, state correlation coefficients, environmental correlation coefficients, and topology correlation coefficients corresponding to the control effect data with the control effect label "poor" under each strategy label are integrated to obtain the third correlation vector. The first and second correlation vectors are used to determine the set of positive correlation features for each strategy label of each topology type; the third correlation vector is used to determine the set of negative correlation features for each strategy label of each topology type.

[0049] (13) (14) (15) (16) The set of positively correlated features is determined as shown in Formulas 13 to 16, where, The weighted correlation coefficient of the b-th feature in the first and second correlation vectors of the j-th strategy label of the i-th topology type label; and These are the weights of the correlation coefficients in the first correlation vector and the weights of the correlation coefficients in the second correlation vector, respectively. ; The correlation coefficient between the b-th feature and the regulation evaluation value in the first correlation vector of the j-th strategy label of the i-th topology type label; The correlation coefficient between the b-th feature and the regulation evaluation value in the second correlation vector of the j-th strategy label of the i-th topology type label; This represents the number of correlation coefficients in the first correlation vector; The feature-related sequence of the j-th strategy label for the i-th topology type label; This indicates that all weighted correlation coefficients of the j-th strategy label for the i-th topology type label are sorted in descending order; The c-th weighted correlation coefficient in the feature correlation sequence of the j-th strategy label of the i-th topology type label; Let be the set of positively correlated features of the j-th strategy label for the i-th topology type label; The feature corresponding to the c-th weighted correlation coefficient in the feature correlation sequence of the j-th strategy label of the i-th topology type label can be source load feature, state feature, environment feature and topology feature. and It can be the source-load correlation coefficient, state correlation coefficient, environmental correlation coefficient, and topological correlation coefficient.

[0050] Sort all correlation coefficients less than 0 in the third correlation vector from largest to smallest. Find the smallest N5 among all the sorted correlation coefficients, such that the sum of the first N5 correlation coefficients is greater than 80% of the sum of the remaining correlation coefficients. Based on the features corresponding to all the smallest N5 correlation coefficients, determine the set of negative correlation features for each strategy label of each topology type label.

[0051] This embodiment achieves refined feature selection by determining the set of positive and negative correlation features.

[0052] In another embodiment of the power distribution network operation control method provided by the present invention, step 400 specifically includes: Step 410: Using the change features as model input, the historical control strategies and control evaluation values ​​corresponding to each topological change in the control effect data as model output, and the relevant feature set as model constraints, construct the operation strategy migration model for each topological type label; Step 420: Integrate the operation strategy migration models of each topology type label to obtain the power grid strategy migration model; Step 430: Regulate the distribution network based on the power grid strategy migration model.

[0053] Specifically, the control effect data corresponding to all control effect labels under each strategy label of each topology type label is extracted. From this control effect data, source load change data, operational status data, and topology feature data for all topology changes are extracted, and these three types of data are used as inputs to the operational strategy transfer model for each topology type label. Historical control strategies and control evaluation values ​​for all topology changes are extracted from this control effect data, and these two types of data are used as outputs to the operational strategy transfer model for each topology type label. The positive and negative correlation feature sets corresponding to each strategy label of each topology type label are extracted, and these two feature sets are used as the core constraints for constructing the operational strategy transfer model for each topology type label. Based on the model training and model construction process, the operational strategy transfer model for each topology type label is constructed sequentially.

[0054] The operation strategy migration models corresponding to all topology type labels are globally integrated to obtain a power grid strategy migration model that can adapt to various topology scenarios of the distribution network. The power grid strategy migration model is used to efficiently regulate the distribution network.

[0055] This embodiment improves the efficiency of the power grid strategy migration model in regulating different topology scenarios of the distribution network by integrating the operation strategy migration models corresponding to the topology type labels of each topology.

[0056] The distribution network operation and control system provided by the present invention is described below. The distribution network operation and control system described below can be referred to in correspondence with the distribution network operation and control method described above.

[0057] Please refer to Figure 3 The present invention also provides a power distribution network operation control system, comprising: The type change data determination module 301 is used to divide the historical topology change data of the distribution network and obtain the type change data of each topology type label; The strategy clustering set determination module 302 is used to cluster the type change data based on the historical control strategies corresponding to each topology type label to obtain a strategy clustering set with strategy labels. The relevant feature set determination module 303 is used to determine the relevant feature set of the strategy label based on the regulation evaluation value corresponding to the strategy cluster set; The distribution network operation and control module 304 is used to control the distribution network based on the historical control strategy and the relevant feature set.

[0058] Optionally, the type change data determination module includes: The label extraction unit is used to extract labels from the topology change data in the historical topology change data to obtain the label values ​​of each branch label. The adjacency matrix construction unit is used to construct an initial adjacency matrix and reconstruct the adjacency matrix based on the label values ​​of each branch label; The topology feature data determination unit is used to extract features from the topology difference graph determined based on the initial adjacency matrix and the reconstructed adjacency matrix to obtain topology feature data. The type change data determination unit is used to divide the historical topology change data based on the topology feature data to obtain the type change data of each topology type label.

[0059] Optionally, the strategy clustering set determination module includes: The historical control strategy determination unit is used to extract the historical strategy corresponding to each topology change in the type change data of each topology type label, and obtain the historical control strategy corresponding to each topology type label. The data clustering unit is used to cluster the type change data based on the control logic and control method of the historical control strategy to obtain the strategy cluster set corresponding to each topology type label; The strategy label determination unit is used to determine the strategy label of the strategy cluster set based on the common characteristics of the strategies in the strategy cluster set.

[0060] Optionally, the relevant feature set determination module includes: The effect evaluation sequence determination unit is used to sort the control evaluation values ​​corresponding to each topological change in the strategy cluster set to obtain the effect evaluation sequence of the strategy label; The comparison result determination unit is used to obtain the mean of the regulation evaluation distribution based on the effect evaluation sequence, and to determine the comparison result between the mean of the regulation evaluation distribution and the mean threshold. The regulation effect label interval determination unit is used to determine the regulation effect label interval of the strategy label based on the effect evaluation sequence and the comparison results; The relevant feature set determination unit is used to determine the relevant feature set of the strategy label based on the regulation evaluation value and the regulation effect label range.

[0061] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for operation and control of a power distribution network, characterized in that, include: The historical topology change data of the distribution network is divided to obtain the type change data of each topology type label; Based on the historical control strategies corresponding to each of the topology type labels, the type change data are clustered to obtain a set of strategy clusters with strategy labels. Based on the regulation evaluation value corresponding to the strategy cluster set, determine the relevant feature set of the strategy label; The power distribution network is regulated based on the historical control strategies and the relevant feature set.

2. The power distribution network operation and control method as described in claim 1, characterized in that, The process of dividing the historical topology change data of the distribution network to obtain type change data for each topology type label includes: Labels are extracted from the topology change data in the historical topology change data to obtain the label values ​​of each branch. Based on the label values ​​of each branch label, an initial adjacency matrix and a reconstructed adjacency matrix are constructed; Feature extraction is performed on the topological difference graph determined based on the initial adjacency matrix and the reconstructed adjacency matrix to obtain topological feature data; Based on the topological feature data, the historical topological change data is divided to obtain the type change data of each topological type label.

3. The power distribution network operation and control method as described in claim 1, characterized in that, The clustering of the type change data based on the historical control strategies corresponding to each of the topology type labels to obtain a strategy cluster set with strategy labels includes: Extract the historical strategies corresponding to each topology change from the type change data of each topology type label to obtain the historical control strategies corresponding to each topology type label; Based on the control logic and control method of the historical control strategy, the type change data are clustered to obtain the strategy cluster set corresponding to each topology type label; Based on the common characteristics of the strategies in the strategy cluster set, the strategy labels of the strategy cluster set are determined.

4. The power distribution network operation and control method as described in claim 1, characterized in that, The step of determining the relevant feature set of the strategy label based on the regulation evaluation value corresponding to the strategy cluster set includes: The regulation evaluation values ​​corresponding to each topological change in the strategy cluster set are sorted to obtain the effect evaluation sequence of the strategy label; Based on the effect evaluation sequence, the mean of the regulation evaluation distribution is obtained, and the comparison result between the mean of the regulation evaluation distribution and the mean threshold is determined. Based on the effect evaluation sequence and the comparison results, the range of the regulatory effect label of the strategy label is determined; Based on the regulation evaluation value and the regulation effect label range, the relevant feature set of the strategy label is determined.

5. The power distribution network operation and control method as described in claim 4, characterized in that, The set of relevant features for determining the strategy label based on the regulation evaluation value and the regulation effect label range includes: Based on the regulation evaluation value and the regulation effect label range, determine the regulation effect data of the strategy label; Determine the change characteristics corresponding to each topological change in the regulation effect data; A correlation analysis is performed on the change characteristics and the regulation evaluation value to obtain the relevant feature set of the strategy label.

6. The power distribution network operation and control method as described in claim 5, characterized in that, The regulation of the distribution network based on the historical regulation strategy and the relevant feature set includes: Using the change features as model input, the historical control strategies and control evaluation values ​​corresponding to each topological change in the control effect data as model output, and the relevant feature set as model constraints, a migration model for the operating strategy of each topological type label is constructed. The operational strategy migration models of each of the aforementioned topology type labels are integrated to obtain the power grid strategy migration model; The power distribution network is regulated based on the power grid strategy migration model.

7. A power distribution network operation and control system, characterized in that, include: The type change data determination module is used to divide the historical topology change data of the distribution network and obtain the type change data of each topology type label; The strategy cluster set determination module is used to cluster the type change data based on the historical control strategies corresponding to each topology type label to obtain a strategy cluster set with strategy labels. The relevant feature set determination module is used to determine the relevant feature set of the strategy label based on the regulation evaluation value corresponding to the strategy cluster set; The distribution network operation and control module is used to control the distribution network based on the historical control strategy and the relevant feature set.

8. The power distribution network operation and control system as described in claim 7, characterized in that, The type change data determination module includes: The label extraction unit is used to extract labels from the topology change data in the historical topology change data to obtain the label values ​​of each branch label. The adjacency matrix construction unit is used to construct an initial adjacency matrix and reconstruct the adjacency matrix based on the label values ​​of each branch label; The topology feature data determination unit is used to extract features from the topology difference graph determined based on the initial adjacency matrix and the reconstructed adjacency matrix to obtain topology feature data. The type change data determination unit is used to divide the historical topology change data based on the topology feature data to obtain the type change data of each topology type label.

9. The power distribution network operation and control system as described in claim 7, characterized in that, The strategy cluster set determination module includes: The historical control strategy determination unit is used to extract the historical strategy corresponding to each topology change in the type change data of each topology type label, and obtain the historical control strategy corresponding to each topology type label. The data clustering unit is used to cluster the type change data based on the control logic and control method of the historical control strategy to obtain the strategy cluster set corresponding to each topology type label; The strategy label determination unit is used to determine the strategy label of the strategy cluster set based on the common characteristics of the strategies in the strategy cluster set.

10. The power distribution network operation and control system as described in claim 7, characterized in that, The relevant feature set determination module includes: The effect evaluation sequence determination unit is used to sort the control evaluation values ​​corresponding to each topological change in the strategy cluster set to obtain the effect evaluation sequence of the strategy label; The comparison result determination unit is used to obtain the mean of the regulation evaluation distribution based on the effect evaluation sequence, and to determine the comparison result between the mean of the regulation evaluation distribution and the mean threshold. The regulation effect label interval determination unit is used to determine the regulation effect label interval of the strategy label based on the effect evaluation sequence and the comparison results; The relevant feature set determination unit is used to determine the relevant feature set of the strategy label based on the regulation evaluation value and the regulation effect label range.

Images